基于PageRank改进的文献价值排序算法

PageRank-based Improved Literature Value Ranking Algorithm

已有学者通过单一指标或者综合因素做了大量文献价值排序工作,但是新旧论文的价值评估不公平现象依旧存在.为了解决这一问题,本文提出一种改进的文献价值排序算法,该算法结合了经典PageRank思想以及常用的被引量与下载量,同时将时间因子引入阻尼因子(d),使得新旧文献在价值评估时有不同的侧重,提高了价值评估公平性.实证分析结果表明,与传统的排序结果相比,新算法的排序更加符合搜寻者的需求.

Scholars have done a lot of the ranking of literature values work through a single indicator or a combination of factors,but the inequality of the evaluating between new and old papers still exists.In order to optimize the ranking problem of the literatures,this paper proposes an improved literature value ranking algorithm,which combines the classic PageRank idea with the commonly used citation and download volume,and introduces the time factor into the damping coefficient(d)to make the new and old literatures have different emphasis in value evaluation,which improves the fairness.The empirical analysis shows that the ranking of the new algorithm is more in line with the needs of the searcher than the traditional sorting results.